open-consul/agent/proxycfg/connect_proxy.go

442 lines
16 KiB
Go

package proxycfg
import (
"context"
"fmt"
"strings"
cachetype "github.com/hashicorp/consul/agent/cache-types"
"github.com/hashicorp/consul/agent/structs"
"github.com/hashicorp/consul/proto/pbpeering"
)
type handlerConnectProxy struct {
handlerState
}
// initialize sets up the watches needed based on current proxy registration
// state.
func (s *handlerConnectProxy) initialize(ctx context.Context) (ConfigSnapshot, error) {
snap := newConfigSnapshotFromServiceInstance(s.serviceInstance, s.stateConfig)
snap.ConnectProxy.DiscoveryChain = make(map[UpstreamID]*structs.CompiledDiscoveryChain)
snap.ConnectProxy.WatchedDiscoveryChains = make(map[UpstreamID]context.CancelFunc)
snap.ConnectProxy.WatchedUpstreams = make(map[UpstreamID]map[string]context.CancelFunc)
snap.ConnectProxy.WatchedUpstreamEndpoints = make(map[UpstreamID]map[string]structs.CheckServiceNodes)
snap.ConnectProxy.WatchedUpstreamPeerTrustBundles = make(map[string]context.CancelFunc)
snap.ConnectProxy.UpstreamPeerTrustBundles = make(map[string]*pbpeering.PeeringTrustBundle)
snap.ConnectProxy.WatchedGateways = make(map[UpstreamID]map[string]context.CancelFunc)
snap.ConnectProxy.WatchedGatewayEndpoints = make(map[UpstreamID]map[string]structs.CheckServiceNodes)
snap.ConnectProxy.WatchedServiceChecks = make(map[structs.ServiceID][]structs.CheckType)
snap.ConnectProxy.PreparedQueryEndpoints = make(map[UpstreamID]structs.CheckServiceNodes)
snap.ConnectProxy.UpstreamConfig = make(map[UpstreamID]*structs.Upstream)
snap.ConnectProxy.PassthroughUpstreams = make(map[UpstreamID]map[string]map[string]struct{})
snap.ConnectProxy.PassthroughIndices = make(map[string]indexedTarget)
snap.ConnectProxy.PeerUpstreamEndpoints = make(map[UpstreamID]structs.CheckServiceNodes)
snap.ConnectProxy.PeerUpstreamEndpointsUseHostnames = make(map[UpstreamID]struct{})
// Watch for root changes
err := s.dataSources.CARoots.Notify(ctx, &structs.DCSpecificRequest{
Datacenter: s.source.Datacenter,
QueryOptions: structs.QueryOptions{Token: s.token},
Source: *s.source,
}, rootsWatchID, s.ch)
if err != nil {
return snap, err
}
err = s.dataSources.TrustBundleList.Notify(ctx, &pbpeering.TrustBundleListByServiceRequest{
// TODO(peering): Pass ACL token
ServiceName: s.proxyCfg.DestinationServiceName,
Namespace: s.proxyID.NamespaceOrDefault(),
Partition: s.proxyID.PartitionOrDefault(),
}, peeringTrustBundlesWatchID, s.ch)
if err != nil {
return snap, err
}
// Watch the leaf cert
err = s.dataSources.LeafCertificate.Notify(ctx, &cachetype.ConnectCALeafRequest{
Datacenter: s.source.Datacenter,
Token: s.token,
Service: s.proxyCfg.DestinationServiceName,
EnterpriseMeta: s.proxyID.EnterpriseMeta,
}, leafWatchID, s.ch)
if err != nil {
return snap, err
}
// Watch for intention updates
err = s.dataSources.Intentions.Notify(ctx, &structs.ServiceSpecificRequest{
Datacenter: s.source.Datacenter,
QueryOptions: structs.QueryOptions{Token: s.token},
EnterpriseMeta: s.proxyID.EnterpriseMeta,
ServiceName: s.proxyCfg.DestinationServiceName,
}, intentionsWatchID, s.ch)
if err != nil {
return snap, err
}
// Get information about the entire service mesh.
err = s.dataSources.ConfigEntry.Notify(ctx, &structs.ConfigEntryQuery{
Kind: structs.MeshConfig,
Name: structs.MeshConfigMesh,
Datacenter: s.source.Datacenter,
QueryOptions: structs.QueryOptions{Token: s.token},
EnterpriseMeta: *structs.DefaultEnterpriseMetaInPartition(s.proxyID.PartitionOrDefault()),
}, meshConfigEntryID, s.ch)
if err != nil {
return snap, err
}
// Watch for service check updates
err = s.dataSources.HTTPChecks.Notify(ctx, &cachetype.ServiceHTTPChecksRequest{
ServiceID: s.proxyCfg.DestinationServiceID,
EnterpriseMeta: s.proxyID.EnterpriseMeta,
}, svcChecksWatchIDPrefix+structs.ServiceIDString(s.proxyCfg.DestinationServiceID, &s.proxyID.EnterpriseMeta), s.ch)
if err != nil {
return snap, err
}
if s.proxyCfg.Mode == structs.ProxyModeTransparent {
// When in transparent proxy we will infer upstreams from intentions with this source
err := s.dataSources.IntentionUpstreams.Notify(ctx, &structs.ServiceSpecificRequest{
Datacenter: s.source.Datacenter,
QueryOptions: structs.QueryOptions{Token: s.token},
ServiceName: s.proxyCfg.DestinationServiceName,
EnterpriseMeta: s.proxyID.EnterpriseMeta,
}, intentionUpstreamsID, s.ch)
if err != nil {
return snap, err
}
}
// Watch for updates to service endpoints for all upstreams
for i := range s.proxyCfg.Upstreams {
u := s.proxyCfg.Upstreams[i]
uid := NewUpstreamID(&u)
// Store defaults keyed under wildcard so they can be applied to centrally configured upstreams
if u.DestinationName == structs.WildcardSpecifier {
snap.ConnectProxy.UpstreamConfig[uid] = &u
continue
}
snap.ConnectProxy.UpstreamConfig[uid] = &u
// This can be true if the upstream is a synthetic entry populated from centralized upstream config.
// Watches should not be created for them.
if u.CentrallyConfigured {
continue
}
dc := s.source.Datacenter
if u.Datacenter != "" {
dc = u.Datacenter
}
if s.proxyCfg.Mode == structs.ProxyModeTransparent && (dc == "" || dc == s.source.Datacenter) {
// In transparent proxy mode, watches for upstreams in the local DC are handled by the IntentionUpstreams watch.
continue
}
// Default the partition and namespace to the namespace of this proxy service.
partition := s.proxyID.PartitionOrDefault()
if u.DestinationPartition != "" {
partition = u.DestinationPartition
}
ns := s.proxyID.NamespaceOrDefault()
if u.DestinationNamespace != "" {
ns = u.DestinationNamespace
}
cfg, err := parseReducedUpstreamConfig(u.Config)
if err != nil {
// Don't hard fail on a config typo, just warn. We'll fall back on
// the plain discovery chain if there is an error so it's safe to
// continue.
s.logger.Warn("failed to parse upstream config",
"upstream", uid.String(),
"error", err,
)
}
switch u.DestinationType {
case structs.UpstreamDestTypePreparedQuery:
err = s.dataSources.PreparedQuery.Notify(ctx, &structs.PreparedQueryExecuteRequest{
Datacenter: dc,
QueryOptions: structs.QueryOptions{Token: s.token, MaxAge: defaultPreparedQueryPollInterval},
QueryIDOrName: u.DestinationName,
Connect: true,
Source: *s.source,
}, "upstream:"+uid.String(), s.ch)
if err != nil {
return snap, err
}
case structs.UpstreamDestTypeService:
fallthrough
case "":
if u.DestinationPeer != "" {
// NOTE: An upstream that points to a peer by definition will
// only ever watch a single catalog query, so a map key of just
// "UID" is sufficient to cover the peer data watches here.
s.logger.Trace("initializing watch of peered upstream", "upstream", uid)
// TODO(peering): We'll need to track a CancelFunc for this
// once the tproxy support lands.
err := s.dataSources.Health.Notify(ctx, &structs.ServiceSpecificRequest{
PeerName: uid.Peer,
Datacenter: dc,
QueryOptions: structs.QueryOptions{
Token: s.token,
},
ServiceName: u.DestinationName,
Connect: true,
// Note that Identifier doesn't type-prefix for service any more as it's
// the default and makes metrics and other things much cleaner. It's
// simpler for us if we have the type to make things unambiguous.
Source: *s.source,
EnterpriseMeta: uid.EnterpriseMeta,
}, upstreamPeerWatchIDPrefix+uid.String(), s.ch)
if err != nil {
return snap, err
}
// Check whether a watch for this peer exists to avoid duplicates.
if _, ok := snap.ConnectProxy.WatchedUpstreamPeerTrustBundles[uid.Peer]; !ok {
peerCtx, cancel := context.WithCancel(ctx)
if err := s.dataSources.TrustBundle.Notify(peerCtx, &pbpeering.TrustBundleReadRequest{
Name: uid.Peer,
Partition: uid.PartitionOrDefault(),
}, peerTrustBundleIDPrefix+uid.Peer, s.ch); err != nil {
cancel()
return snap, fmt.Errorf("error while watching trust bundle for peer %q: %w", uid.Peer, err)
}
snap.ConnectProxy.WatchedUpstreamPeerTrustBundles[uid.Peer] = cancel
}
continue
}
err = s.dataSources.CompiledDiscoveryChain.Notify(ctx, &structs.DiscoveryChainRequest{
Datacenter: s.source.Datacenter,
QueryOptions: structs.QueryOptions{Token: s.token},
Name: u.DestinationName,
EvaluateInDatacenter: dc,
EvaluateInNamespace: ns,
EvaluateInPartition: partition,
OverrideMeshGateway: s.proxyCfg.MeshGateway.OverlayWith(u.MeshGateway),
OverrideProtocol: cfg.Protocol,
OverrideConnectTimeout: cfg.ConnectTimeout(),
}, "discovery-chain:"+uid.String(), s.ch)
if err != nil {
return snap, fmt.Errorf("failed to watch discovery chain for %s: %v", uid.String(), err)
}
default:
return snap, fmt.Errorf("unknown upstream type: %q", u.DestinationType)
}
}
return snap, nil
}
func (s *handlerConnectProxy) handleUpdate(ctx context.Context, u UpdateEvent, snap *ConfigSnapshot) error {
if u.Err != nil {
return fmt.Errorf("error filling agent cache: %v", u.Err)
}
switch {
case u.CorrelationID == rootsWatchID:
roots, ok := u.Result.(*structs.IndexedCARoots)
if !ok {
return fmt.Errorf("invalid type for response: %T", u.Result)
}
snap.Roots = roots
case strings.HasPrefix(u.CorrelationID, peerTrustBundleIDPrefix):
resp, ok := u.Result.(*pbpeering.TrustBundleReadResponse)
if !ok {
return fmt.Errorf("invalid type for response: %T", u.Result)
}
peer := strings.TrimPrefix(u.CorrelationID, peerTrustBundleIDPrefix)
if resp.Bundle != nil {
snap.ConnectProxy.UpstreamPeerTrustBundles[peer] = resp.Bundle
}
case u.CorrelationID == peeringTrustBundlesWatchID:
resp, ok := u.Result.(*pbpeering.TrustBundleListByServiceResponse)
if !ok {
return fmt.Errorf("invalid type for response: %T", u.Result)
}
if len(resp.Bundles) > 0 {
snap.ConnectProxy.InboundPeerTrustBundles = resp.Bundles
}
snap.ConnectProxy.InboundPeerTrustBundlesSet = true
case u.CorrelationID == intentionsWatchID:
resp, ok := u.Result.(structs.Intentions)
if !ok {
return fmt.Errorf("invalid type for response: %T", u.Result)
}
snap.ConnectProxy.Intentions = resp
snap.ConnectProxy.IntentionsSet = true
case u.CorrelationID == intentionUpstreamsID:
resp, ok := u.Result.(*structs.IndexedServiceList)
if !ok {
return fmt.Errorf("invalid type for response %T", u.Result)
}
seenUpstreams := make(map[UpstreamID]struct{})
for _, svc := range resp.Services {
uid := NewUpstreamIDFromServiceName(svc)
seenUpstreams[uid] = struct{}{}
cfgMap := make(map[string]interface{})
u, ok := snap.ConnectProxy.UpstreamConfig[uid]
if ok {
cfgMap = u.Config
} else {
// Use the centralized upstream defaults if they exist and there isn't specific configuration for this upstream
// This is only relevant to upstreams from intentions because for explicit upstreams the defaulting is handled
// by the ResolveServiceConfig endpoint.
wildcardSID := structs.NewServiceID(structs.WildcardSpecifier, s.proxyID.WithWildcardNamespace())
wildcardUID := NewUpstreamIDFromServiceID(wildcardSID)
defaults, ok := snap.ConnectProxy.UpstreamConfig[wildcardUID]
if ok {
u = defaults
cfgMap = defaults.Config
snap.ConnectProxy.UpstreamConfig[uid] = defaults
}
}
cfg, err := parseReducedUpstreamConfig(cfgMap)
if err != nil {
// Don't hard fail on a config typo, just warn. We'll fall back on
// the plain discovery chain if there is an error so it's safe to
// continue.
s.logger.Warn("failed to parse upstream config",
"upstream", uid,
"error", err,
)
}
meshGateway := s.proxyCfg.MeshGateway
if u != nil {
meshGateway = meshGateway.OverlayWith(u.MeshGateway)
}
watchOpts := discoveryChainWatchOpts{
id: NewUpstreamIDFromServiceName(svc),
name: svc.Name,
namespace: svc.NamespaceOrDefault(),
partition: svc.PartitionOrDefault(),
datacenter: s.source.Datacenter,
cfg: cfg,
meshGateway: meshGateway,
}
up := &handlerUpstreams{handlerState: s.handlerState}
err = up.watchDiscoveryChain(ctx, snap, watchOpts)
if err != nil {
return fmt.Errorf("failed to watch discovery chain for %s: %v", uid, err)
}
}
snap.ConnectProxy.IntentionUpstreams = seenUpstreams
// Clean up data from services that were not in the update
for uid, targets := range snap.ConnectProxy.WatchedUpstreams {
if upstream, ok := snap.ConnectProxy.UpstreamConfig[uid]; ok && upstream.Datacenter != "" && upstream.Datacenter != s.source.Datacenter {
continue
}
if _, ok := seenUpstreams[uid]; !ok {
for _, cancelFn := range targets {
cancelFn()
}
delete(snap.ConnectProxy.WatchedUpstreams, uid)
}
}
for uid := range snap.ConnectProxy.WatchedUpstreamEndpoints {
if upstream, ok := snap.ConnectProxy.UpstreamConfig[uid]; ok && upstream.Datacenter != "" && upstream.Datacenter != s.source.Datacenter {
continue
}
if _, ok := seenUpstreams[uid]; !ok {
delete(snap.ConnectProxy.WatchedUpstreamEndpoints, uid)
}
}
for uid, cancelMap := range snap.ConnectProxy.WatchedGateways {
if upstream, ok := snap.ConnectProxy.UpstreamConfig[uid]; ok && upstream.Datacenter != "" && upstream.Datacenter != s.source.Datacenter {
continue
}
if _, ok := seenUpstreams[uid]; !ok {
for _, cancelFn := range cancelMap {
cancelFn()
}
delete(snap.ConnectProxy.WatchedGateways, uid)
}
}
for uid := range snap.ConnectProxy.WatchedGatewayEndpoints {
if upstream, ok := snap.ConnectProxy.UpstreamConfig[uid]; ok && upstream.Datacenter != "" && upstream.Datacenter != s.source.Datacenter {
continue
}
if _, ok := seenUpstreams[uid]; !ok {
delete(snap.ConnectProxy.WatchedGatewayEndpoints, uid)
}
}
for uid, cancelFn := range snap.ConnectProxy.WatchedDiscoveryChains {
if upstream, ok := snap.ConnectProxy.UpstreamConfig[uid]; ok && upstream.Datacenter != "" && upstream.Datacenter != s.source.Datacenter {
continue
}
if _, ok := seenUpstreams[uid]; !ok {
cancelFn()
delete(snap.ConnectProxy.WatchedDiscoveryChains, uid)
}
}
for uid := range snap.ConnectProxy.PassthroughUpstreams {
if _, ok := seenUpstreams[uid]; !ok {
delete(snap.ConnectProxy.PassthroughUpstreams, uid)
}
}
for addr, indexed := range snap.ConnectProxy.PassthroughIndices {
if _, ok := seenUpstreams[indexed.upstreamID]; !ok {
delete(snap.ConnectProxy.PassthroughIndices, addr)
}
}
// These entries are intentionally handled separately from the WatchedDiscoveryChains above.
// There have been situations where a discovery watch was cancelled, then fired.
// That update event then re-populated the DiscoveryChain map entry, which wouldn't get cleaned up
// since there was no known watch for it.
for uid := range snap.ConnectProxy.DiscoveryChain {
if upstream, ok := snap.ConnectProxy.UpstreamConfig[uid]; ok && upstream.Datacenter != "" && upstream.Datacenter != s.source.Datacenter {
continue
}
if _, ok := seenUpstreams[uid]; !ok {
delete(snap.ConnectProxy.DiscoveryChain, uid)
}
}
case strings.HasPrefix(u.CorrelationID, "upstream:"+preparedQueryIDPrefix):
resp, ok := u.Result.(*structs.PreparedQueryExecuteResponse)
if !ok {
return fmt.Errorf("invalid type for response: %T", u.Result)
}
pq := strings.TrimPrefix(u.CorrelationID, "upstream:")
uid := UpstreamIDFromString(pq)
snap.ConnectProxy.PreparedQueryEndpoints[uid] = resp.Nodes
case strings.HasPrefix(u.CorrelationID, svcChecksWatchIDPrefix):
resp, ok := u.Result.([]structs.CheckType)
if !ok {
return fmt.Errorf("invalid type for service checks response: %T, want: []structs.CheckType", u.Result)
}
svcID := structs.ServiceIDFromString(strings.TrimPrefix(u.CorrelationID, svcChecksWatchIDPrefix))
snap.ConnectProxy.WatchedServiceChecks[svcID] = resp
default:
return (*handlerUpstreams)(s).handleUpdateUpstreams(ctx, u, snap)
}
return nil
}